B American Society for Mass Spectrometry, 2015
J. Am. Soc. Mass Spectrom. (2015) DOI: 10.1007/s13361-015-1124-z
RESEARCH ARTICLE
Extensive Charge Reduction and Dissociation of Intact Protein Complexes Following Electron Transfer on a Quadrupole-Ion Mobility-Time-of-Flight MS Frederik Lermyte,1,2 Jonathan P. Williams,3 Jeffery M. Brown,3 Esther M. Martin,1 Frank Sobott1,2 1
Biomolecular and Analytical Mass Spectrometry group, Department of Chemistry, University of Antwerp, Groenenborgerlaan 171, 2020, Antwerp, Belgium 2 Center for Proteomics (CFP-CeProMa), University of Antwerp, Antwerp, Belgium 3 Waters Corporation, Stamford Avenue, Altrincham Road, Wilmslow, SK9 4AX, UK
Abstract. Non-dissociative charge reduction, typically considered to be an unwanted side reaction in electron transfer dissociation (ETD) experiments, can be enhanced significantly in order to reduce the charge state of intact protein complexes to as low as 1+ on a commercially available Q-IM-TOF instrument. This allows for the detection of large complexes beyond 100,000 m/z, while at the same time generating top-down ETD fragments, which provide sequence information from surface-exposed parts of the folded structure. Optimization of the supplemental activation has proven to be crucial in these experiments and the charge-reduced species are most likely the product of both proton transfer (PTR) and non-dissociative electron transfer (ETnoD) reactions that occur prior to the ion mobility cell. Applications of this approach range from deconvolution of complex spectra to the manipulation of charge states of gas-phase ions. Keywords: Top-down fragmentation, Electron transfer dissociation, Noncovalent complex, Charge state reduction, Protein complex Received: 26 September 2014/Revised: 25 February 2015/Accepted: 1 March 2015
Introduction
M
ass spectrometry (MS) has the ability to provide both molecular mass and sequence information of biomolecules, following their ionization [1, 2]. In electrospray ionization (ESI), proteins typically exhibit multiple charges that result from protonation of solvent accessible sites. The number of charges (z) can be high, either because the proteins are unfolded, or as a consequence of their large mass and size. The charge state distribution of an ion is an important indicator of its threedimensional structure in the gas phase, and for proteins it is generally believed to be closely linked to their folding state [3]. Under nondenaturing conditions (Bnative MS^), nanoelectrospray ionization is commonly used. DC and rf-voltages
Electronic supplementary material The online version of this article (doi:10.1007/s13361-015-1124-z) contains supplementary material, which is available to authorized users. Correspondence to: Frank Sobott; e-mail: [email protected]
together with instrumental gas pressures are tuned such that weak biomolecular interactions are preserved and maintained during analysis. Native MS involves the ionization, transmission, and detection of intact, noncovalently assembled protein– protein, protein–DNA/RNA, and protein–ligand complexes [2, 4, 5]. Spectra typically contain relatively narrow charge state distributions at comparatively high m/z. However, because of the presence of potentially overlapping charge states and significant peak widths, charge state assignment can still be challenging, particularly for polydisperse samples or in complex mixtures. As a result, a number of specialized algorithms have been proposed to deconvolute these spectra and identify the components [6–12]. Another approach is to minimize mass spectral complexity by reducing the charge state of the ions in the gas phase by means of Bcharge-stripping^ [13–17]. In these experiments, charge reduction and the subsequent transmission and detection of the resulting extremely high m/z species typically require instrument modifications [15, 18]. Limited charge reduction is also typically observed as a side reaction in electron capture (ECD) and electron transfer dissociation (ETD) experiments, without instrument modification,
F. Lermyte et al.: Charge Reduction of Intact Protein Complexes
particularly when working with intact proteins [19–24]. In addition, ECD or ETD fragments can remain bound to each other by noncovalent interactions so that they appear as chargereduced precursor ions in the spectra [24–27]. In some cases, this limited charge reduction can be sufficient to resolve complex spectra where mass and charge state assignment are otherwise not trivial [28]. Here we describe a new approach to manipulate the ion charge state using unmodified, ETDenabled QTOF type instruments (Synapt G2 and G2-S, Waters Corporation, Wilmslow, UK) with standard ETD reagents [29–32], and propose the term Bcharge-reduction ETD^ or short BcrETD^ for it, in analogy to Bcharge reduction electrospray mass spectrometry^ (CREMS) introduced by Smith and colleagues [13, 14]. Using this methodology, we are able to observe ions with a reduced net positive charge as low as 1+ from native protein complexes, whilst also generating product ions that yield sequence-specific information within the same experiment. Importantly, the ability to manipulate charge states in the gas phase over such a wide range also opens up new possibilities for the investigation of electrostatic effects on the folding state of protein ions in the gas phase, independent of how they were generated, and to minimize the effect of charge repulsion on native structures [33, 34]. An interesting study in this regard was recently carried out by Campuzano and Schnier, who used a corona discharge to induce significant charge reduction of peptides and proteins on a Synapt HDMS instrument [35, 36]. However, as with earlier charge reduction studies mentioned above, this required instrument modification. Also, the observed charge reduction was limited, as no detection of signals above approximately 18,000 m/z was reported. Furthermore, as charge reduction occurred in the source region of the instrument, a specific precursor could not be selected by using the quadrupole mass filter of the instrument in this case, an issue that is addressed by the approach described here.
aqueous ammonium acetate. All experiments were carried out on Synapt G2 and G2-S HDMS mass spectrometers (Waters Corporation, Wilmslow, UK) in TOF mode, unless indicated otherwise. Approximately 5 μL of protein solution was transferred to an in-house prepared gold-coated capillary and infused into the mass spectrometer using the nanoflow version of the Z-spray ion source, using a capillary voltage of 1.0–1.3 kV, minimal nanoflow gas pressure, and a backing pressure of 5 mbar. The implementation of ETD on Synapt instruments has been described in detail elsewhere [30] and tuning of the glow discharge was as described earlier [24]. The time windows during which protein cations and ETD reagent anions were accumulated in the trap cell were 1.0 s and 0.1 s, respectively. Traditionally, specific proton transfer reagents such as perfluoro-1,3-dimethylcyclohexane (PDCH) are used for charge-reduction studies [15]. However, standard ETD reagents were used for these experiments: 1,4-dicyanobenzene on the Synapt G2 (fitted with a 32 k quadrupole), and pnitrotoluene on the Synapt G2-S (8 k quadrupole). Use of either reagent led to similar spectra and showed more efficient charge reduction than PDCH (data not shown). Instrument settings were as follows on both Synapt platforms: trap pressure 8e-2 mbar, trap collision energy 10 V, trap DC bias 11 V (45 V in Mobility mode), transfer pressure 6e-3 mbar (2e-2 mbar in Mobility mode), transfer collision energy 5 V, detector voltage 3000 V, sampling cone 60–120 V, trap wave height 0.5–0.7 V (1.6–1.9 V in Mobility mode) (sampling cone voltage and trap wave height optimized for each protein complex). For the ion mobility experiments, the IM wave height was 30 V, IM wave velocity 1000 m/s, and the pressure in the IM cell was 4.3 mbar (He gas flow 180 mL/min; IM cell gas flow 50 mL/min).
Results and Discussion Experimental Five proteins, all of which self-assemble into larger oligomers in their native state, were chosen to demonstrate extensive charge reduction by ETD: hemoglobin (Bos taurus, Sigma H2500, 64 kDa), concanavalin a (Canavalia ensiformis, Sigma C2010, 103 kDa), alcohol dehydrogenase (Saccharomyces cerevisiae, Sigma A3263, 148 kDa), pyruvate kinase (Oryctolagus cuniculus, Sigma P9136, 233 kDa), and Lglutamic dehydrogenase (Bos taurus, Sigma G7882, 336 kDa). L-glutamic dehydrogenase is hexameric, whereas the other four proteins form tetramers. The lyophilized proteins were dissolved at a concentration of 1 mg/mL in 100 mM aqueous ammonium acetate (pH=6.8) and desalted twice using Bio-Rad Micro Bio-Spin 6 columns, yielding an estimated final concentration of the complex between 3 and 11 μM. The polydisperse αB-crystallin (Homo sapiens, monomer mass 20.2 kDa) used to demonstrate separation of overlapping peaks was a gift from the Laboratory for Biocrystallography at KU Leuven, Belgium, and was used at a concentration of 1.3 mg/mL, also in 100 mM
The mass spectra displayed on the left in Figure 1 show the charge state distributions of native hemoglobin, concanavalin a, alcohol dehydrogenase, pyruvate kinase, and L-glutamic dehydrogenase, acquired in positive ion mode without the introduction of ETD reagent anions into the instrument (glow discharge emitter switched off). The spectra shown in Figure 1a were acquired on a Synapt G2 (up to 100,000 m/z), and those in Figure 1b on a Synapt G2-S instrument (up to 120,000 m/z). The charge state that was selected in the quadrupole for charge reduction is indicated in red. The result of ETD conditions, which promote extensive charge reduction of the complexes, is shown on the right-hand side of Figure 1. Interestingly, a bi- or trimodal charge state distribution is typically observed, with a first intensity maximum in close proximity to the selected precursor ions, whereas the others are located at much lower charge states. The gradual decrease in signal intensity above 50,000 m/z is most likely due to less efficient ion transmission and reduced sensitivity of the detector in the high m/z range. Another striking feature of the spectra in Figure 1 is the peak broadening observed with increasing m/z. It should be noted,
F. Lermyte et al.: Charge Reduction of Intact Protein Complexes
Figure 1. Left column: native charge state distributions of hemoglobin, alcohol dehydrogenase, L-glutamic dehydrogenase, concanavalin a, and pyruvate kinase with the precursor charge state in red. Right column: product ion spectra with chargereduced complex signals. The spectra in (a) were acquired on the Synapt G2 in the mass range 3000–100,000 m/z, whereas the spectra in (b) were acquired on the G2-S, from 3000–120,000 m/z
however, that the actual resolution of the peaks, defined as (m/z)/(Δm/z), either remains constant (e.g., in the case of Lglutamate dehydrogenase) or decreases by a factor of less than two (e.g., for hemoglobin). Another possible reason for apparent peak broadening is the formation of adducts with the ETD reagent [37, 38]. As the mass of the anion (128 Da for the 1,4dicyanobenzene used in this study) is between 0.04% and 0.2% of the mass of the complexes studied here, the resulting adduct peaks would not be resolved in the spectra shown in Figure 1. The observed peak widths are, however, only compatible with a few adducts. Whereas the charge-reduced tetramers are annotated as Qn+ for the sake of simplicity, it should be noted that these are not the same [Q+nH]n+ charge states produced by electrospray ionization, but rather the products of a combination of proton transfer and nondissociative electron transfer (ETnoD), which are both known to occur under these conditions [24, 37–40]. It
would, therefore, be more accurate to consider them as [Q+xH]n+, with n
J. Am. Soc. Mass Spectrom. (2015) DOI: 10.1007/s13361-015-1124-z
RESEARCH ARTICLE
Extensive Charge Reduction and Dissociation of Intact Protein Complexes Following Electron Transfer on a Quadrupole-Ion Mobility-Time-of-Flight MS Frederik Lermyte,1,2 Jonathan P. Williams,3 Jeffery M. Brown,3 Esther M. Martin,1 Frank Sobott1,2 1
Biomolecular and Analytical Mass Spectrometry group, Department of Chemistry, University of Antwerp, Groenenborgerlaan 171, 2020, Antwerp, Belgium 2 Center for Proteomics (CFP-CeProMa), University of Antwerp, Antwerp, Belgium 3 Waters Corporation, Stamford Avenue, Altrincham Road, Wilmslow, SK9 4AX, UK
Abstract. Non-dissociative charge reduction, typically considered to be an unwanted side reaction in electron transfer dissociation (ETD) experiments, can be enhanced significantly in order to reduce the charge state of intact protein complexes to as low as 1+ on a commercially available Q-IM-TOF instrument. This allows for the detection of large complexes beyond 100,000 m/z, while at the same time generating top-down ETD fragments, which provide sequence information from surface-exposed parts of the folded structure. Optimization of the supplemental activation has proven to be crucial in these experiments and the charge-reduced species are most likely the product of both proton transfer (PTR) and non-dissociative electron transfer (ETnoD) reactions that occur prior to the ion mobility cell. Applications of this approach range from deconvolution of complex spectra to the manipulation of charge states of gas-phase ions. Keywords: Top-down fragmentation, Electron transfer dissociation, Noncovalent complex, Charge state reduction, Protein complex Received: 26 September 2014/Revised: 25 February 2015/Accepted: 1 March 2015
Introduction
M
ass spectrometry (MS) has the ability to provide both molecular mass and sequence information of biomolecules, following their ionization [1, 2]. In electrospray ionization (ESI), proteins typically exhibit multiple charges that result from protonation of solvent accessible sites. The number of charges (z) can be high, either because the proteins are unfolded, or as a consequence of their large mass and size. The charge state distribution of an ion is an important indicator of its threedimensional structure in the gas phase, and for proteins it is generally believed to be closely linked to their folding state [3]. Under nondenaturing conditions (Bnative MS^), nanoelectrospray ionization is commonly used. DC and rf-voltages
Electronic supplementary material The online version of this article (doi:10.1007/s13361-015-1124-z) contains supplementary material, which is available to authorized users. Correspondence to: Frank Sobott; e-mail: [email protected]
together with instrumental gas pressures are tuned such that weak biomolecular interactions are preserved and maintained during analysis. Native MS involves the ionization, transmission, and detection of intact, noncovalently assembled protein– protein, protein–DNA/RNA, and protein–ligand complexes [2, 4, 5]. Spectra typically contain relatively narrow charge state distributions at comparatively high m/z. However, because of the presence of potentially overlapping charge states and significant peak widths, charge state assignment can still be challenging, particularly for polydisperse samples or in complex mixtures. As a result, a number of specialized algorithms have been proposed to deconvolute these spectra and identify the components [6–12]. Another approach is to minimize mass spectral complexity by reducing the charge state of the ions in the gas phase by means of Bcharge-stripping^ [13–17]. In these experiments, charge reduction and the subsequent transmission and detection of the resulting extremely high m/z species typically require instrument modifications [15, 18]. Limited charge reduction is also typically observed as a side reaction in electron capture (ECD) and electron transfer dissociation (ETD) experiments, without instrument modification,
F. Lermyte et al.: Charge Reduction of Intact Protein Complexes
particularly when working with intact proteins [19–24]. In addition, ECD or ETD fragments can remain bound to each other by noncovalent interactions so that they appear as chargereduced precursor ions in the spectra [24–27]. In some cases, this limited charge reduction can be sufficient to resolve complex spectra where mass and charge state assignment are otherwise not trivial [28]. Here we describe a new approach to manipulate the ion charge state using unmodified, ETDenabled QTOF type instruments (Synapt G2 and G2-S, Waters Corporation, Wilmslow, UK) with standard ETD reagents [29–32], and propose the term Bcharge-reduction ETD^ or short BcrETD^ for it, in analogy to Bcharge reduction electrospray mass spectrometry^ (CREMS) introduced by Smith and colleagues [13, 14]. Using this methodology, we are able to observe ions with a reduced net positive charge as low as 1+ from native protein complexes, whilst also generating product ions that yield sequence-specific information within the same experiment. Importantly, the ability to manipulate charge states in the gas phase over such a wide range also opens up new possibilities for the investigation of electrostatic effects on the folding state of protein ions in the gas phase, independent of how they were generated, and to minimize the effect of charge repulsion on native structures [33, 34]. An interesting study in this regard was recently carried out by Campuzano and Schnier, who used a corona discharge to induce significant charge reduction of peptides and proteins on a Synapt HDMS instrument [35, 36]. However, as with earlier charge reduction studies mentioned above, this required instrument modification. Also, the observed charge reduction was limited, as no detection of signals above approximately 18,000 m/z was reported. Furthermore, as charge reduction occurred in the source region of the instrument, a specific precursor could not be selected by using the quadrupole mass filter of the instrument in this case, an issue that is addressed by the approach described here.
aqueous ammonium acetate. All experiments were carried out on Synapt G2 and G2-S HDMS mass spectrometers (Waters Corporation, Wilmslow, UK) in TOF mode, unless indicated otherwise. Approximately 5 μL of protein solution was transferred to an in-house prepared gold-coated capillary and infused into the mass spectrometer using the nanoflow version of the Z-spray ion source, using a capillary voltage of 1.0–1.3 kV, minimal nanoflow gas pressure, and a backing pressure of 5 mbar. The implementation of ETD on Synapt instruments has been described in detail elsewhere [30] and tuning of the glow discharge was as described earlier [24]. The time windows during which protein cations and ETD reagent anions were accumulated in the trap cell were 1.0 s and 0.1 s, respectively. Traditionally, specific proton transfer reagents such as perfluoro-1,3-dimethylcyclohexane (PDCH) are used for charge-reduction studies [15]. However, standard ETD reagents were used for these experiments: 1,4-dicyanobenzene on the Synapt G2 (fitted with a 32 k quadrupole), and pnitrotoluene on the Synapt G2-S (8 k quadrupole). Use of either reagent led to similar spectra and showed more efficient charge reduction than PDCH (data not shown). Instrument settings were as follows on both Synapt platforms: trap pressure 8e-2 mbar, trap collision energy 10 V, trap DC bias 11 V (45 V in Mobility mode), transfer pressure 6e-3 mbar (2e-2 mbar in Mobility mode), transfer collision energy 5 V, detector voltage 3000 V, sampling cone 60–120 V, trap wave height 0.5–0.7 V (1.6–1.9 V in Mobility mode) (sampling cone voltage and trap wave height optimized for each protein complex). For the ion mobility experiments, the IM wave height was 30 V, IM wave velocity 1000 m/s, and the pressure in the IM cell was 4.3 mbar (He gas flow 180 mL/min; IM cell gas flow 50 mL/min).
Results and Discussion Experimental Five proteins, all of which self-assemble into larger oligomers in their native state, were chosen to demonstrate extensive charge reduction by ETD: hemoglobin (Bos taurus, Sigma H2500, 64 kDa), concanavalin a (Canavalia ensiformis, Sigma C2010, 103 kDa), alcohol dehydrogenase (Saccharomyces cerevisiae, Sigma A3263, 148 kDa), pyruvate kinase (Oryctolagus cuniculus, Sigma P9136, 233 kDa), and Lglutamic dehydrogenase (Bos taurus, Sigma G7882, 336 kDa). L-glutamic dehydrogenase is hexameric, whereas the other four proteins form tetramers. The lyophilized proteins were dissolved at a concentration of 1 mg/mL in 100 mM aqueous ammonium acetate (pH=6.8) and desalted twice using Bio-Rad Micro Bio-Spin 6 columns, yielding an estimated final concentration of the complex between 3 and 11 μM. The polydisperse αB-crystallin (Homo sapiens, monomer mass 20.2 kDa) used to demonstrate separation of overlapping peaks was a gift from the Laboratory for Biocrystallography at KU Leuven, Belgium, and was used at a concentration of 1.3 mg/mL, also in 100 mM
The mass spectra displayed on the left in Figure 1 show the charge state distributions of native hemoglobin, concanavalin a, alcohol dehydrogenase, pyruvate kinase, and L-glutamic dehydrogenase, acquired in positive ion mode without the introduction of ETD reagent anions into the instrument (glow discharge emitter switched off). The spectra shown in Figure 1a were acquired on a Synapt G2 (up to 100,000 m/z), and those in Figure 1b on a Synapt G2-S instrument (up to 120,000 m/z). The charge state that was selected in the quadrupole for charge reduction is indicated in red. The result of ETD conditions, which promote extensive charge reduction of the complexes, is shown on the right-hand side of Figure 1. Interestingly, a bi- or trimodal charge state distribution is typically observed, with a first intensity maximum in close proximity to the selected precursor ions, whereas the others are located at much lower charge states. The gradual decrease in signal intensity above 50,000 m/z is most likely due to less efficient ion transmission and reduced sensitivity of the detector in the high m/z range. Another striking feature of the spectra in Figure 1 is the peak broadening observed with increasing m/z. It should be noted,
F. Lermyte et al.: Charge Reduction of Intact Protein Complexes
Figure 1. Left column: native charge state distributions of hemoglobin, alcohol dehydrogenase, L-glutamic dehydrogenase, concanavalin a, and pyruvate kinase with the precursor charge state in red. Right column: product ion spectra with chargereduced complex signals. The spectra in (a) were acquired on the Synapt G2 in the mass range 3000–100,000 m/z, whereas the spectra in (b) were acquired on the G2-S, from 3000–120,000 m/z
however, that the actual resolution of the peaks, defined as (m/z)/(Δm/z), either remains constant (e.g., in the case of Lglutamate dehydrogenase) or decreases by a factor of less than two (e.g., for hemoglobin). Another possible reason for apparent peak broadening is the formation of adducts with the ETD reagent [37, 38]. As the mass of the anion (128 Da for the 1,4dicyanobenzene used in this study) is between 0.04% and 0.2% of the mass of the complexes studied here, the resulting adduct peaks would not be resolved in the spectra shown in Figure 1. The observed peak widths are, however, only compatible with a few adducts. Whereas the charge-reduced tetramers are annotated as Qn+ for the sake of simplicity, it should be noted that these are not the same [Q+nH]n+ charge states produced by electrospray ionization, but rather the products of a combination of proton transfer and nondissociative electron transfer (ETnoD), which are both known to occur under these conditions [24, 37–40]. It
would, therefore, be more accurate to consider them as [Q+xH]n+, with n
